Causal and Fair Machine Learning
The exact topic depends on the student's interest, student's background, and previous research experience. Generally speaking, there are mainly three topics of this project:
1. Designing Fair Machine Learning Algorithms. In this project, students will focus on how to make the current machine learning algorithms be fair. They will also explore the fairness issue of the current machine learning algorithms, especially for healthcare data.
2. Causality as a tool for de-biasing current deep learning algorithms. Students will use the idea of causality to different deep learning tasks to de-bias the datasets or algorithms in order to improve the accuracy and trustworthiness.
3. Causality as a tool for invariant learning. This project mainly focuses on transfer learning, students will use causality to design transfer learning algorithms.
Computer, Electrical and Mathematical Sciences and Engineering
Field of Study -
Causal Inference, Fairness, Transfer Learning, Deep Learning
Desired Project Deliverables
During the project, students will have opportunity to learn about some topics in trustworthy machine learning, especially fair learning, transfer learning and causal learning. They will learn and implement the SOTA methods. Hopefully, they may produce some publication after the intern.